|
Evaluating
Heterodox Theories*
Evan Fales and Barry
Markovsky
The University of Iowa
Mail questions and observations for Professors Fales and Markovsky to trance@nowhere.nowhere. Their gathered answers will be published here on an as needed basis.
Abstract
Active and heterogeneous disciplines
constantly spawn new theories and theoretical variants. By
definition, each such offering is heterodox to the degree that
its veracity would diminish accepted theories. Most often
heterodox theories are dismissed out of hand for non-rational
reasons, e.g., they just seem too bizarre. Most of the time, too,
rational analysis supports such rejection. Of course, many
important theories in science once seemed bizarre but later were
accepted as evidence accumulated for them and against received
views. The lag between a premature rejection and ultimate
acceptance is an inefficiency built into the theory evaluation
process. Is there a way to reduce this inefficiency? Through
examining a heterodox sociological exemplar, we discuss the
standards to which such theories should be held in order to
deserve (1) hearings in their relevant disciplines, (2) serious
attention, and (3) assignment of a high likelihood of being true.
In every scientific discipline there arise from
time to time challenges to "mainstream thinking" that
appeal to heterodox theoretical frameworks or observational
claims, or both. How should such challenges be treated? On the
one hand, novelty is the lever of progress, opening to view new
possibilities. On the other hand, no discipline can afford to
devote serious attention to every unorthodox notion that comes
over the horizon. This problem is an especially sensitive one for
sociology, given the complexity and multi-level nature of the
phenomena we study, and our correspondingly rudimentary
understanding of them.
It is not our purpose to say how heterodox
theories are in fact produced or received by scientific
disciplines. Instead, we address normative issues: (1) the
standards to which a heterodox theory should be held in order to
merit a hearing; (2) the criteria it must satisfy to merit
serious attention within a discipline; and (3) methods for
determining its likelihood of being true. We provide general
criteria for theory evaluation, then discuss special problems of,
and guidelines for, the heterodox. Our thesis is illustrated by
an analysis of a published report claiming that people practicing
Transcendental Meditation (TM) at one geographic location have
immediate beneficial effects on social indicators at distant
locations.
Evaluating Theories
Scientific theories (henceforward
"theories") consist of structured sets of claims,
subject to evaluation via objective criteria (Cohen 1989;
Markovsky 1994, 1996). These criteria bear upon a theory’s
internal structure, its standing vis-à-vis prior theories, and
its relationship to the empirical world.
Theory Structure
Theoretical arguments consist of statements
that consist of terms. Because scientific theories offer unique,
non-intuitive ways of understanding the world, they frequently
employ terms with unfamiliar meanings. To be communicable,
however, all terms must be understood in the same way by members
of the intended audience. This requires a hierarchical conceptual
system. At its base are undefined or primitive terms whose
meanings are shared by the theorist and audience. Primitive terms
are crafted into combinations which compose definitions or
indicate correct usage for some defined terms. In turn,
higher-order concepts are conveyed by definitions that include
primitives and/or terms previously defined. A parsimonious
conceptual system imparts clear meanings to just those terms
needed to express the theory’s statements.
A theory may be unimpeachable formally, but if
meanings of terms are unclear, deriving and interpreting
predictions becomes a haphazard affair. A theory purporting to
have scientific credentials must be articulated with sufficient
precision and specificity that the contents of its claims can be
linked to empirical referents and procedures for testing. That
is, it must present a sufficiently precise picture of the
constituents and causal processes of some natural domain that one
can construct procedures for detecting and measuring the
properties of those constituents. Those procedures also will
employ causal knowledge concerning the ways that measuring
devices or empirical phenomena should be influenced by the
mechanisms under investigation.
Vagueness about a theory’s empirical
referents or how they interact with previously understood parts
or our world will undermine any attempt to claim that certain
phenomena provide evidence for (or against) that theory. For to
say that some observed phenomenon measures, or is sensitive to,
the presence of some underlying mechanism presupposes some idea
of how that mechanism behaves—and specifically, how it can
be expected to behave in the conditions under which the
observation was made.
Even if all its terms are well-defined, a
theory still can be toppled by a single logical flaw. If one
claims empirical confirmation for a hypothesis derived from a
theory, but logical analysis finds that the statement is the
product of an invalid argument, then no longer can the
theory justify the prediction, connect to the hypothesis, or
benefit from the test results.
Confirmation Issues
There is no fully agreed-upon metric for theory
confirmation. However, a Bayesian framework captures
several universal scientific values: For given levels of
empirical evidence and prior knowledge, a theory’s
confirmation is greater to the extent that (1) it is compelling
in view of prior knowledge; (2) the datum whose evidential value
is to be assessed is made probable by theory and prior knowledge;
and (3) the datum is not probable relative to alternative
hypotheses and prior knowledge.[1] Even without precise measures of these components,
Bayes' Theorem permits ceteris paribus assessments of
relative confirmation across theories. So with all else
approximately equal between Theories A and B, if A explains the
evidence better than B, then A has the higher confirmation status
of the two. Furthermore, when all three criteria favor B over A,
no ceteris paribus provision is needed: B will always have
the higher confirmation status.
Special Considerations for the Heterodox
Even if analysis reveals terminological
ambiguities or questionable logic, publication still may be
warranted if a theory is the best in its class. However, journals
have limited space and cannot print every such effort, so the
decision process becomes more complicated. Should an exciting,
new, but untested theory have priority over a relatively workaday
confirmation of a well-established formulation? What of the
heterodox theory that comes bolstered by evidence, but flies in
the face of established knowledge?
There are good reasons to publish heterodox
work. First, it is difficult to justify rejecting a submission on
grounds that it is heterodox when otherwise it satisfies normal
standards. Second, heterodox views, if put forward with some
competence, deserve a hearing if only because sometimes one of
them turns out to be right and makes a major advance. This
mandates a policy of relative tolerance in publication decisions.
The low prior probability of a theory should not contribute, at
least in a direct way, to its being denied a hearing.
On the other hand, there are good reasons to
subject heterodox viewpoints to higher than normal standards: A
corollary of Bayes’ theorem suggests that data providing
significant and exclusive support for a very improbable theory
deserve especially cautious and careful scrutiny[2] Whatever evidence we have for the improbability of a
heterodox theory is, a fortiori, evidence for the
illegitimacy of data purportedly favoring that theory. The case
against the heterodox is strongest when a competing hypothesis
can be established.
Is it right, however, that orthodoxy must
supply a detailed rebuttal of all the data in order to deflect
every heterodox challenge? It depends. For instance, if "all
the data" are ten similarly flawed experiments, then
refuting one of them refutes the other nine. Or if the claimant
asserts that one study supplies especially critical support, then
the identification of flaws in that study provides an equally
critical falsification. In fact, it is sufficient that a
well-entrenched position offer plausible alternative
explanations for the challenging data. The burden of
refuting those explanations lies with the proponents of the
heterodox challenge.[3]
The Theory Behind the
"Maharishi Effect"
To illustrate our thesis, we examine a study
that was organized by researchers affiliated with the
Transcendental Meditation (TM) movement founded by Maharishi
Mahesh Yogi. Orme-Johnson, Alexander, Davies, Chandler and
Larimore (1988, henceforward O88) purportedly confirmed a theory
asserting that variations in the number of
TM-Sidhis—advanced TM practitioners—in a given location
cause measurable variations in social phenomena elsewhere through
the operation of a "unified field." The research was
conducted in East Jerusalem and employed social indicators from
Jerusalem, Israel, and Lebanon. Meditators’ participation
was voluntary and self-selected, their numbers varying from 65 -
241 between Aug. 1 - Sept. 30, 1983. The investigators predicted
and found that meditation group size correlated negatively with
crime rates, automobile accidents, and fires in Jerusalem;
positively with stock market prices and a national mood indicator
in Israel; and negatively with war intensity and war fatalities
in Lebanon.
Researchers, spokespersons, and politicians
affiliated with the TM movement claim that this so-called
"Maharishi Effect" (ME) is now scientifically proven
through rigorous empirical research, replicated numerous times,
and reported in peer-reviewed journals. Newspaper accounts,
promotional materials, subsequent research reports, and our own
communications with TM researchers, representatives, followers
and defectors all indicate that, perhaps more than any other, the
article we shall discuss provides a special source of pride,
vindication, and scientific legitimation for all affiliated with
the movement. In the sub-section to follow we examine the
coherence of the theory behind the ME and its standing vis-à-vis
prior theories. Following that we address research-related
issues, noting methodological loopholes and alternative
explanations.
The Theory
O88 devoted considerable space to the Maharishi
Effect theory (MT). Its central ideas come from Maharishi’s
theory of consciousness, developed further by physicist John
Hagelin (1987, 1989).[4]
Following are O88’s core argument and key terms, with
numbers added for subsequent reference.
[1] collective consciousness is the
wholeness of consciousness of the group, that is more than
the sum of the consciousnesses of all individuals composing
that group. [p.778]
[2] [The theory] posits a unified field of
"pure consciousness" as the basis of the diverse
activities of all individual minds. All processes of thought
and perception are viewed as fluctuations or qualified
expressions of this underlying, unqualified, least-excited
state of consciousness. Maharishi likens the individual mind
transcending its more active levels and experiencing its
basis in pure consciousness to a localized wave settling to
become the silent, unbounded surface of the ocean. Such
experiences are said to create nonlocal, fieldlike effects of
order and coherence in the environment... . [p.778-9]
[3] nonlocal effects could be mediated
through the agency of the unified quantum field due to the
intrinsically nonlocal structure of space-time at this scale.
[p.784]
[4] A potential explanation for the
apparent propagation of such coherent effects may lie,
however, at the ultimate scale of superunification, the
Planck scale of 10-33 cm and 10-44 sec, where the fundamental forces and matter fields
are said to become fully unified... [p.784]
[5] the localized conscious awareness of
the individual becomes experientially connected back to pure
consciousness, the unified source of order and intelligence
at its base, thus increasing coherence, reducing stress, and
accelerating development in the larger society. [p.784]
TM thus is assumed to permit the individual to
experience his/her minds’ "basis in pure
consciousness," which is embedded in the "unified
quantum field" at extremely small scales [4].[5] Individual and environment are linked by this field
[3], and so the consciousnesses of meditators cause
"nonlocal effects" [2] or actions-at-a-distance.
Meditators acting in concert create, it is claimed, a greater
wave of coherence in the unified quantum field than could be
achieved by meditators acting separately. The ME influences
anyone in the field’s reach [5], creating in them TM-like
coherence. The affected population is  , where N1 is the number of
meditators in the population, N2 is the TM-Sidhi group
size, a and b are approximately 100, and N2
> 100.
Theory Analysis
The theory receives low marks for
meaningfulness. Key terms are undefined or only roughly
characterized using other complex, undefined terms or metaphors.
"Planck scale" and "unified quantum field"
are defined in physics, but the meanings of many crucial
expressions are not so clear, including
"consciousness," "collective consciousness"
(CC), "pure consciousness," and "experientially
connected." Consider CC, defined loosely in [1]. First, it
relies on another undefined term (consciousness), the meaning of
which is not self-evident. Second, the expression "the
wholeness of" is vacuous. Third, identifying what CC is
more than still does not tell the reader what it is.
Linking CC to "pure consciousness" does not help
either, for the latter is characterized through an avalanche of
still more vague esoterica. Without clear definitions, the
authors must rely on metaphors—an ocean surface, a laser,
radio signals—all of which break down. Unlike the ocean
surface, the laser and radio signals, CC is, respectively,
non-material, omni-directional, and in violation of the
inverse-square law of signal strength [p.785].
To evaluate the plausibility of the argument,
readers (and journal referees) would need solid grounding in
contemporary physics. We consulted several nuclear and particle
physicists and learned that detailed experimental evidence is
lacking for Planck scale phenomena. Also, a number of unified
theories compatible with the existing experimental constraints
have been considered in this highly speculative area (Davies and
Brown 1988).[6] Moreover, physicists examining purported links to MT
find them highly dubious (e.g., Stenger 1990; Pagels 1986). Thus,
although O88 give the impression that their assumptions are
well-grounded, the soundness of MT’s quantum field
connections is an open question at best.
Though more rigorous than the rest of the
theory, the ME equation also has problems. First, no rationale is
offered for its thresholds. The cut-offs of 100 are arbitrary,
and the functional discontinuity they entail produces a rather
awkward behavioral model. Second, why is the measurement unit number
of people rather than, say, physical distance? The
implication is paradoxical. Assume there is a 100-person TM-Sidhi
group in downtown Chicago and another in rural Fairfield, Iowa.
Ignoring for this example the smaller effect of non-TM-Sidhi
meditators, ME = 1,000,000 for both groups. This means that the
effect would have a radius of possibly a couple of miles for the
Chicago group, but more than 50 times that distance for the other
group. Chicago’s unified field then would somehow have to
"know" that it is not supposed to affect people if they
happen to be more than 27 (or however many) blocks from the
meditation group, and Fairfield’s unified field would have
to realize that it must keep going (and going) to affect its
quota of citizens. Although O88 claim many things for their
unified field, this level of sentience is not one of them.[7]
Theory Context
One component of Bayes’ theorem pertains
to theoretical contexts—the likelihood that the theory is
correct in view of its consistency with, and plausibility
relative to, prior knowledge. In statements such as [3] above,
O88 imply that MT is consistent with the theories of contemporary
physics. Indeed, Hagelin (1987, 1989) argues first that the
identification of the mental with the physical is plausible
within the framework of quantum mechanics and, second, that this
identification—and a fully worked-out theory of mind/matter
that anticipates contemporary unified-field theories—is
found in the oldest sacred Hindu texts, the Vedas.
Hagelin’s case rests on three pillars: (1)
supernormal phenomena ("sidhis"), such as the ME,
levitation and invisibility, have no other natural explanation;
(2) parallels between an esoteric theory of consciousness
and quantum mechanics; and (3) parallels between a theory
allegedly imbedded in the Vedas and contemporary unified-field
theories. Regarding (1), Hagelin goes to some lengths to invoke
the more recherché possibilities allowable by quantum theory, in
particular, levitation. The problem is that, ignoring for now the
ME, no sidhis are validated and so the far-fetched explanation
lacks purpose. Under the circumstances, Hagelin assigns premier
importance to the ME, saying it provides "the central core
of experimental evidence in support of the proposed identity
between pure consciousness and the unified field" (1987:73).
Hagelin’s (and O88’s) other pillars
are equally shaky. His argument for a unified field-consciousness
identity suggests that some quantum-mechanical properties of
physical fields match characteristics of consciousness. However,
his argument relies critically upon ambiguity and obscurity in
the terms denoting these properties. For instance, he notes that
"creativity of consciousness" describes intellectual
inventiveness, whereas "creativity of matter" describes
the quantum field’s capacity to generate particles. Both
kinds of creativity share the characteristic of production,
but Hagelin does nothing to show that these two kinds of
production are the same, or even interestingly analogous. This is
about as cogent as arguing that the mind is a sort of mirror
because both reflect, but it does capture the essence of
Hagelin’s approach.
To draw his parallels between the Vedas and
contemporary unified-field theories, Hagelin relies on
numerological and exegetic styles of reasoning. For example, his
evidence for linkages between the Vedas, the unified quantum
field, and consciousness includes the fact that there are five
special Vedic terms called tanmatras, and there are five "spin-types" in
quantum mechanics. Moreover, the quantum field theory that
Hagelin especially favors is known among physicists as a
"superstring" theory. In corroboration, Hagelin offers
a line from a Vedic text that he translates as "My body is
called a string."[8]
Hagelin’s interpretations of physical fields in terms of
consciousness are supported by nothing more than the construction
of arbitrary formal isomorphisms, metaphors, and a reliance upon
ambiguity and vague analogy.
In sum, O88’s theory does not pass minimal
criteria of meaningfulness and logical integrity. Even if it did,
where the theory in question is that which is at best
foreshadowed in Hagelin’s writings, and prior knowledge
includes knowledge of physical laws, neurophysiology and the
like, the probability of the MT in view of prior knowledge is
very close to zero.[9]
Evidence for the
"Maharishi Effect"
Even if the MT is not compelling, the kinds of
empirical results claimed to support the theory would be
noteworthy if true. As indicated by Abelson (forthcoming)[10] and Schrodt (1990), however, the empirical claims are
weakened considerably by the nature of the research design and
statistical techniques used by O88. These critics emphasize
problems endemic to non-random treatment conditions, and the
proneness of time-series analyses to Type II (false positive)
errors.[11] In contrast, we take a more direct approach suggested
by the two remaining components of Bayes’ theorem: whether
the evidence offered for the MT is actually predicted by the
theory, and whether specific alternative hypotheses explain the
evidence.[12]
Do Predictions Derive From the Theory?
To be credible, MT must explain (1) how group
meditation affects the unified field, (2) how these effects in
turn cause changes in the actions of individual human beings, and
(3) how those individual actions have their claimed social
impact. Not clear in the theory is how meditators’ brains
spontaneously synchronize to produce "coherence," how
coherence modifies physio-chemical structures within the brains
of distant others, why coherence only causes phenomena that
happen to be defined socio-culturally as "positive,"
and how the distribution of behaviors is affected at the
population level.[13]
Hagelin (1987:69) concedes that MT does not
explain how meditators affect the ostensible unified field, and
the emotions, thoughts, and behaviors of people within it.
Moreover, if the material world is presumed to be influenced by
collective meditation (as it must be if the unified field is
affected), then more direct measures are possible: inanimate ME-detectors
should be placed at varying distances from the meditation group.[14] As it stands,
causal linkages from individual TM practitioners to, for
instance, a diminished likelihood of Nebraskans wrecking their
cars, are hidden in a very "black box."
Another indication that predictions do not
derive from the theory is that lagged effects were not specified
prior to the research. The researchers then interpret any
significant correlation at any non-negative time lag for any
indicator as supporting their theory. This multiplies the
likelihood of finding "supportive" evidence, but opens
the door to Type II errors even wider.
The ME equation, from which specific hypotheses
might have been derived, was ignored completely in the research.
The time-series analysis employed each day’s higher number
of Sidhi meditators rather than its square, and the number of
non-Sidhi meditators was not included in the test. There also
were periods during the study when meditation group size fell
below assumed thresholds for affecting Lebanon, or both Lebanon
and Israel, but Sidhi-group size rather than zero was used as the
independent variable.[15] In
fact, the Sidhi group should not have been powerful enough to
influence the Lebanese war, or even most of Israel. O88 argue
that the roughly 38,000 non-Sidhi meditators in Israel and 2,000
more in Lebanon provided the needed boost. However, O88 provide
no information about the geographic distribution of these
meditators or evidence to support their assumption that they were
in the area and meditating during the vacation month of August
which comprises half the test period.
In sum, ME predictions cannot be derived from
the MT. There are gaps in the causal chain from group meditation
to the phenomena supposedly affected, there are no specified time
lags for the ME, and despite the capacity of MT’s formal
component to generate specific ME predictions, the model is
ignored. Thus, evidence offered for the ME cannot significantly
increase confidence in the veracity of the MT.
Can Alternative Hypotheses Explain the
Evidence?
O88 presented their empirical findings several
ways, but the most compelling showed superimposed graphs of (1) a
daily composite index of Jerusalem crime rates, auto accidents
and fires; Israeli crimes and "national mood," and
Lebanese war intensity; and (2) the number of meditators in the
larger of each day’s two meditation groups. Time-series
analyses confirmed a statistically significant relationship
between the number of meditators and the composite index for lags
3 0.
Our analysis focuses on a few specific factors that were not
included in the analysis but readily could have influenced both
the meditators’ decisions to participate and the
social indicators used as dependent variables. Importantly, the
research design prevents us from knowing how many other factors
may have contributed to a spurious correlation.
Holidays and Vacation Month. O88
recognize that mundane events affect both meditation group size
and their social indicators: Their statistical model compensates
for weekend effects and for three Jewish holidays. There are
other factors for which the authors failed to invoke statistical
controls, however. Perhaps the most obvious omission was that of
three other major Jewish holidays during the study
period—Succoth, Shemini Azeret, and Simhat Torah. Nearly all
of the meditators who participated in the research came from the
liberal end of the Jewish religious spectrum.[16] As was clear in the graph of meditation group sizes,
they were willing to travel and meditate on the Sabbath and
holidays. In contrast, Orthodox Jews do not cook, light fires, or
travel in cars on religious holidays or the Sabbath. Undoubtedly
the holidays also produce some elevation in mood, and may
coincide with reductions in war hostilities measured from nearby
Lebanon. Therefore, the same factors that increase meditation
group sizes at certain times would also correspond to the
abstinence from cooking and travel for a very significant portion
of the Israeli population. We might then expect to find not only
a heightening in national mood at the same time larger numbers of
meditators turned out for the study, but also coincident
reductions in domestic fires and automobile accidents.
Another effect, obvious in the graph of
meditation group sizes, is attributable to August being vacation
month in Israel. During August Israelis leave the country in
large numbers. Critically, the investigators did not take into
account the effects on fires and auto accidents of the lower
population in Israel during August, of the tendency to cook less
during hot weather, and of the potential for reduced war
hostilities in the desert heat.[17] It may be difficult, but the burden of eliminating
these possibilities rests upon the researchers.
The Lebanon War. O88 highlighted the
negative effects of meditation group size on war intensity in
Lebanon. Hostilities in the Lebanon war involved an enormously
complex interaction between a multitude of social, political and
military forces. A good sense of this complexity can be gleaned
from The New York Times Index "Middle East"
entries for this period.[18] We
offer several observations:
O88 make no mention of the many
widely-publicized military and political events that may have
both influenced the Lebanon war and induced meditators’
patterns of participation. Many such events occurred around the
midpoint of the study, coinciding with the wildest fluctuations
in group size and war intensity. For example: (1) Israel
announced it would withdraw its army from the Shouf mountains
overlooking Beirut to a line along a river to the south. Major
fighting erupted in Beirut just before the announced withdrawal
date. (2) During the study period Prime Minister Begin both
announced his intention to resign, and did so. (3) The Lebanese
army completed a successful sweep of Beirut. A lull then occurred
until the Israelis withdrew from the Shouf. Fighting promptly
erupted among various factions. (4) Within days battles were
being waged over the Shouf by U.S. and Druse militia, and the
U.S. congress voted to keep the Marines in Lebanon for 18 more
months. No effort was made to ask meditators why they showed up
or stayed home in droves at various times during this period of
the study.[19]
In sum, it is hardly unreasonable to suppose
that the fluctuations of the dependent variables measured by O88
would have remained exactly as they were even if there had been
no meditators at all. The claim that MT provides the only
plausible explanation of these data cannot be sustained. There
are alternative explanations that do not depend on esoteric or
paranormal influences.
Discussion and Conclusion
We have devoted considerable space to the
analysis of a particular heterodox report, the upshot of which is
that, at this time, the claims of that report do not merit being
taken seriously by the scientific community. The theory
motivating the research is ill-constructed and not compelling in
view of prior knowledge; the evidence offered is not impressive
and mundane alternative hypotheses offer plausible explanations
for the findings. Only if its data were above suspicion, and no
alternative explanations were known, would MT warrant any
significant confidence in its truth. It is only by acquiring
evidence of a sufficient quality and quantity, rendering any
thought of its rejection irrational, that the heterodox theory
ultimately can triumph.
Defenders of O88 might complain that we have
not really explained their data. After all, we have not
demonstrated that the causal factors we cite account for the
observed correlations, nor have we re-analyzed the data to show
that the ME vanishes when controls are entered for all those test
periods when known exogenous factors might be influencing the
results (which is almost for the entire duration of the
experiment). Such an objection would be entirely misplaced. The
ball is in the other court: It is sufficient, for the purposes of
defense, that a well-entrenched position offer plausible alternative
explanations for the challenging data. The detailed task of
ruling out those explanations lies properly with the challengers.
Failing that, the heterodox theory cannot rise even to the level
of being worthy of serious consideration.
Our criticisms may be divided into those
directed against the MT and those disputing interpretations of
their data. As to the first, the main points were that the MT has
serious problems regarding the clarity and integrity of its
arguments, and it does not cohere well with other strongly
confirmed theories, hence conflicting with the evidence
supporting those theories. MT is under-articulated, often vague
or enigmatic, reliant upon specious analyses, and silent about
key processes that link causes to their alleged effects. These
defects are not uncommon in novel theories, but in this case they
allow nothing better than crude plausibility arguments for its
extraordinary predictions. Contrast this with, say, the theory of
relativity, which was not only formulated in a highly precise
fashion, but which entailed relatively simple and numerically
precise predictions for its tests. So MT has a low prior
probability. That, after all, is what makes it unorthodox, and
its road to plausibility is bound to be arduous. Not only must a
great deal of supporting data be amassed but, most likely, an
indefinitely large body of established science will have to be
overturned or revised to accommodate the new results.[20]
This brings us to interpretation of evidence.
The MT predicts correlations that are supposed to support its
causal claims. Against those claims we have launched, in
descending order of importance, (1) arguments that offer concrete
explanations of findings without invoking the ME; (2) more
speculative arguments from plausible serendipity; and (3)
arguments that do not explain a correlation between two
variables, but that suggest the dependent variable can be fully
explained without recourse to Sidhi group sizes. In the first
class fall our suggestions concerning fires, auto accidents,
national mood, and the immediate consequences of the Begin
resignation; in the second category, our speculations about the
lull in the war during part of August; and in the third, our
mention of some of the many factors affecting the vicissitudes of
the war generally. Even without quantitative support, this sort
of defense of normal science is sufficient to undermine claims of
plausibility on behalf of unorthodox theories that claim
quantitative support. The low prior probability of a heterodox
challenger suffices to establish a presumption of guilt.
A further point is relevant to the social and
historical sciences, which often treat events that, once passed,
can never be duplicated fully. Even worse, relevant details,
unless recognized to be significant and recorded at the time,
later may be unrecoverable. In this case normal science,
inevitably a Johnny-come-lately upon the scene, would suffer an
irremediable disadvantage if the demand for a fully-developed
alternative explanation had to be met. Thus, unless a heterodox
challenger can produce experiments that are well-controlled,
thoroughly analyzable, and replicable, the standard of disproof
to which normal science can legitimately be held is quite low.
Even arguments of the weakest class (3, above) must be reckoned
to weigh heavily against any theory with a prior probability as
low as that of MT.
Finally, we do not pretend to resolve two
difficult but important practical issues. First, to what extent
should the scientific community devote time, effort, and journal
space to debates over unconventional theories, at the expense of
more conventional work that is more likely to be fruitful? We
offer two brief observations. First, when research is conducted
on behalf of an organized group (as was O88), its results will
often find publication in arenas that do not afford an
opportunity for informed rebuttal. Second, such rebuttals (as we
have shown) need not be overly expansive to undermine an
unorthodox theory.
As for the second issue, publication confers a
certain aura of legitimacy in the eyes of the lay public and even
the research community. Proponents of unorthodox theories know
this and, as with TM proponents, often attempt to parlay such
recognition into research grants, influence upon public
policy-makers, and influence with the public at large. To what
extent does the scientific community have a responsibility to
allow—or not allow—these considerations to influence
its handling of unconventional proposals? How much responsibility
the scientific community bears is, in part, a function of the
significance of the consequences of public acceptance.
Publication of an unorthodox medical claim, for example, might
clearly have severe consequences, whereas publication of some
novel view about the formation of igneous rocks most likely would
not. It is here, in any case, that a much wider understanding of
the principle that publication should not be taken to confer
respectability in the sense of acceptability, would be
beneficial.
References
Abelson, Robert P. Forthcoming. "Editorial
Comment." Journal of Conflict Resolution.
Aron, Elaine, and Arthur Aron. 1986. The
Maharishi Effect: A Revolution Through Meditation. Walpole,
NH: Stillpoint Publishing.
Cohen, Bernard P. 1989. Developing
Sociological Knowledge. Chicago: Nelson-Hall.
Davies, P.C.W., and J. Brown. Superstrings:
A Theory of Everything? New York: Cambridge.
Fales, Evan, and B. Markovsky. 1995.
"Evaluating Heterodox Theories: The Maharishi’s
Extraordinary Social Physics." Presented at the Annual
Meetings of the American Sociological Association, Washington,
D.C.
Falk, Arthur. 1995. "Wisdom Updated."
Philosophy of Science 62:389-403.
Frank, Benis M. 1987. U.S. Marines in
Lebanon 1982–1984. Washington, D.C.: History and Museums
Division, Headquarters, U.S. Marine Corps. (U.S. Government
Printing Office.)
Friedman, Thomas L. 1991. From Beirut to
Jerusalem (2nd Edition). New York: Farrar, Strauss, Giroux.
Hagelin, John S. 1987. "Is Consciousness
the Unified Field? A Field Theorist’s Perspective." Modern
Science and Vedic Science 1:28-87.
-----. 1989. "Restructuring Physics From
its Foundation in Light of Maharishi Vedic Science." Modern
Science and Vedic Science 3:3-72.
Hatchard, Guy D., Ashley J. Deans, Kenneth L.
Cavanaugh and David Orme-Johnson. 1996. "The Maharishi
Effect: A Model for Social Improvement. Time Series Analysis of a
Phase Transition to Reduced Crime in Merseyside Metropolitan
Area." Psychology, Crime and Law 2:165-174.
Hume, David. 1955[1748]. An Inquiry
Concerning Human Understanding. New York: Liberal Arts Press.
Markovsky, Barry. 1994. "The Structure of
Theories." In Martha Foschi and Edward J. Lawler (eds.), Group
Processes: Sociological Analyses. Chicago: Nelson-Hall.
------. 1996. "Theory, Science, and
"Micro-macro" Bridges in Structural Social
Psychology." Current Research in Social Psychology 1(4):30-42,
http://www.uiowa.edu/~grpproc.
Orme-Johnson, David W., Charles N. Alexander,
John L. Davies, Howard M. Chandler, and Wallace E. Larimore.
1988. "International Peace Project in the Middle East: The
effects of the Maharishi Technology of the Unified Field." Journal
of Conflict Resolution 32:776-812.
Pagels, Heinz R. 1986. Affidavit No. 85-2849,
United States District Court for the District of Columbia.
Radin, Dean I., Jannine M. Rebman, and Maikwe
P. Cross. 1996. "Anomalous Organization of Random Events by
Group Consciousness: Two Exploratory Experiments." Journal
of Scientific Exploration 10(1):143-168.
Schiff, Ze’ev, and Ehud Ya’ari. 1984.
Israel’s Lebanon War. New York: Simon and Schuster.
Schrodt, Philip A. 1990. "A Methodological
Critique of a Test of the Effects of the Maharishi Technology of
the Unified Field." Journal of Conflict Resolution 34:745-55.
Stenger, Victor J. 1990. Physics and
Psychics. Buffalo, NY: Prometheus.
Wallace, Robert Keith. 1989. The
Neurophysiology of Enlightenment. Fairfield, IA: MIU Press.
Yaniv, Avner. 1987. Dilemmas of Security:
Politics, Strategy, and the Israeli Experience in Lebanon. New
York: Oxford University Press.
Notes
[*] Copyright (c) 1997 The
University of North Carolina Press. Published December 1997 in Social
Forces Volume 76 (2):511-25. Reprinted with permission.
The authors contributed equally to
this project. Direct correspondence to Evan Fales, Department of
Philosophy, or Barry Markovsky, Department of Sociology,
University of Iowa, Iowa City, IA, 52242. E-mail
evan-fales@uiowa.edu or barry-markovsky@uiowa.edu. We are
grateful to Scott Eliason, Jeffrey Erger, Donald Krieger, Michael
Lovaglia, and Philip Schrodt for their insights and comments on
an earlier draft. A much more comprehensive treatment of the
issues and analyses in this article is available from the authors
upon request.
[return]
[1] This is expressed most rigorously by the
formula  , where P
is the probability or level of confirmation of the theory (T), e is
the empirical evidence for T, and k is prior
knowledge relevant to T. The formula asserts that T is
confirmed to the degree that P(T|k) ®
1, P(e|T&k) ®
1, and P(e|k) ® 0.
[return]
[2] If P(T|k) »
0, and if, for every competing alternative Ti,
P(e|Ti&k) »
0, then, even if P(e|T&k) is high, P(e|k) will
be very low. This follows from the expansion  This generalizes to the
case where k < 1, where k includes
well-established theories and their supporting data. In effect,
because those data support theories that conflict with T, they
conflict with e. Unless the truth of e has been
established beyond any possibility of doubt, this weakens
rational credence in e. See Falk (1995) for a formal
Bayesian analysis.
[return]
[3] In his classic discussion "Of
Miracles," Hume (1955[1748]) defends an even more cavalier
dismissal of heterodoxy. He observes that it is not incumbent
upon one to uncover fraud or error in dismissing every miracle
report. Nevertheless, we can feel confident in such a dismissal
because the occurrence of the miracle would violate our
well-entrenched understanding of nature, and would therefore be
placed in competition for our allegiance with the enormous mass
of evidence which supports that understanding. It will therefore
be more likely than not that some undetected-perhaps
undetectable-mistake lurks behind the problematic data.
[return]
[4] See also Wallace (1989) for a more
comprehensive but less detailed review. Hagelin is a faculty
member of the Maharishi University of Management and was the
Natural Law Party's presidential candidate in 1992 and 1996. The
NLP promotes TM as the solution to virtually all of society's
ills.
[return]
[5] Aron and Aron (1986:11) state this more
bluntly: "...[T]he pure consciousness experience is actually
the subjective experience of what in physics is called the
'unified field.'"
[return]
[6] We are grateful to Wayne Polyzou, Yannick
Meurice and Bill Klink of the University of Iowa Department of
Physics and Astronomy for their input.
[return]
[7] The intuitive answer is that individuals
in the field "use up" the "energy" generated
by the meditators. However, this is not our understanding of how
the underlying physical field is presumed to operate, nor is it a
process that we have found discussed in the TM literature.
[return]
[8] Frederick Smith (personal communication),
a Sanskritist at the University of Iowa, notes that Hagelin's
translation is an esoteric one, dramatically different from that
which standard Sanskritic scholarship delivers.
[return]
[9] Important general questions have been
raised by philosophers about the content of k, prior
knowledge. For practical purposes and in the present
context, however, it seems entirely fair to include in k, as
we have done, those well-established theories and data from
physics and biology to which Hagelin et al. themselves appeal.
[return]
[10] In 1994 we were informed by the Editor
of the Journal of Conflict Resolution that Abelson's paper
was forthcoming. It has not yet appeared at this writing.
[return]
[11] Schrodt (1990) cites research
demonstrating that business cycles "cause" sunspots and
that eggs "cause" chickens (but not the reverse). The
problem is that the noisier the data, the greater the range of
frequencies it contains. Standard time-series methods then become
ideally suited to extracting whatever "effect" one
desires.
[return]
[12] A re-analysis of the data also would
have been desirable, though not essential for our case. However,
David Orme-Johnson has refused numerous requests for a copy of
the raw data set.
[return]
[13] Beyond the purview of this critique are
the moral and ethical issues that arise when meditators purport
to alter experimentally the moods, thought processes, and
behaviors of others without their informed consent, and when the
TM movement assigns itself the responsibility of manipulating the
substrate of all existence.
[return]
[14] This is the approach adopted by Radin
and colleagues in their consciousness research (e.g., Radin,
Rebman and Cross 1996). They look for attentional effects in the
anomalous output sequences of random number generators.
[return]
[15] We argue below that Sidhi group size is
affected by some of the same factors that influence the social
indicators. Including Sidhi group size in the analysis when it is
below threshold is thus likely to increase the
"confirmatory" observations, artificially increase the
apparent ME, and further bias results in a direction favoring the
hypothesis.
[return]
[16] Confirmed in conversation with David
Orme-Johnson and Charles Alexander.
[return]
[17] Confirmed in conversation with
Alexander.
[return]
[18] We also relied on more extensive
analyses provided by Frank (1987), Friedman (1991), Schiff and
Ya'ari (1984), and Yaniv (1987).
[return]
[19] Confirmed in conversation with
Alexander.
[return]
[20] TM researchers claim over 40
replications of the Maharishi Effect. Of those which are
published, most are in obscure or newer journals (e.g., Hatchard,
Deans, Cavanaugh and Orme-Johnson 1996). Moreover no two
"replications" that we have reviewed actually employ
identical procedures, measures and/or model specifications, and
none of the field studies employed safeguards against the sorts
of problems we raise.
[return]
The truth on Transcendental Meditation
Please send questions, comments, problems, and letters to the editor to trance@nowhere.nowhere. All editorial
correspondence becomes the property of trancenet.org -- unless requested otherwise -- and may be edited for purposes of clarity and space.
Except where noted, entire contents Copyright ©1995-1998 trancenet.org Society.
trancenet.orgTM is a trademark of trancenet.org Society, an unincorporated nonprofit organization.
The opinions and viewpoints of contributors do not necessarily reflect those of trancenet.org, its editorial staff, nor trancenet.org Society, its board, officers, employees, volunteers.
Neither trancenet.org Society nor its editorial staff conclude that any group discussed on this site is necessarily cultic in nature. We provide suppressed and alternative information
and champion your right make informed decisions for yourself.
Copyrighted works are reprinted with permission as noted or are made available under the "fair use" exception of U.S. copyright law, for research and educational purposes only.
To comment on this or any other trancenet.org page, go to trancechat.
This page was last built with Frontier on a Macintosh on Tue, Jul 21, 1998 at 10:59:34 AM .
|
